Kaplan Moll Violante (2018,AER) show in appendix A.5 a decomposition of a consumption response to a monetary policy shock. They decompose its response to direct and indirect effects. They do that using Dynare.
Does anyone know how they manage to do that in practice?
Hey Johannes, unfortunately they have not uploaded the replication codes for the appendix section where they apply this methodology for the Smets and Wouters model which they say clearly that was implemented in Dynare.
Some others suggested me that in order to keep the impact of one variable only (for example the wage) due to a monetary policy shock, you keep all other prices at the steady state apart from the variable of interest and you solve the model again. Nevertheless I am not sure that this is the right way.
I have contacted the authors, they told me that are working on it but they haven’t uploaded the files yet.
The solution that I found reading some other papers is the one I describe above, although I am not sure that this is what KMV do.
Set the rest of the prices in the SS and leave freely only the one you want to get the effect from.